Loading Now

Temasek comfortable with current AI spending levels, executive says

Temasek comfortable with current AI spending levels, executive says

Temasek comfortable with current AI spending levels, executive says


(Corrects paragraph 7 to say Americas accounts for 26% of portfolio, not US accounts)

BOSTON, Oct 1 (Reuters) – Singapore’s Temasek is not worried about the massive levels of spending from hyperscalers powering the AI buildout, its executive said on Thursday, as the state investor works to increase its exposure to the technology.

Speaking at the Reuters Investment USA event in Boston, Temasek’s North America Head Jane Atherton said the companies funding the infrastructure believe the large-scale investments will yield risk-adjusted returns with the growing adoption of AI.

“Underpinning all of this AI buildout are some of the strongest balance sheets in the world… we remain incredibly positive on AI,” Atherton said.

A selloff in global bonds and expectations of tighter monetary policy have called into question the sustainability of the enormous spending commitments undertaken by tech giants in a high-interest rate era.

Analysts have forecast hyperscaler capex to surpass over $1 trillion by 2027, driven predominantly by the need to fund costly chips and data centers to satisfy massive amounts of compute.

Temasek, which owns stakes in Anthropic and OpenAI, is one of the world’s largest state investors with overall holdings valued at about $400 billion.

The firm’s holdings in the Americas account for 26% of the portfolio, predominantly in the US, with significant stakes in companies spanning financial services and technology, counting BlackRock, Mastercard and Nvidia among its major investments.

Temasek laid out plans earlier this year to increase its investment in the AI ecosystem to as much as 15% of the portfolio over the next five years, from about 6% now.

The conversation around the speed of AI development has taken center stage after leading companies have called for a slowdown, warning that rapid advances could risk spelling the end of humanity.

Atherton said the question of AI safety needed to be addressed by governments and the companies developing the technology.

“I really don’t know how you price the end of humanity, except that if it happens, it won’t really matter,” Atherton said.

(Reporting by Paritosh Bansal in Boston and Utkarsh Shetti in Bengaluru; Editing by Arun Koyyur)